Melancholia EEG Classification Based on CSSD and SVM

摘要：

It takes an important role to get the disease information from melancholia electroencephalograph (EEG). Firstly, A common spatial subspace decomposition (CSSD) method was used to extract features from 16-channel EEG of melancholia and normal healthy persons. Then based on support vector machines (SVM), a classifier was designed to train and test its classification capability between Melancholia and healthy persons. The results indicated that the proposed method can reach a higher accuracy as 95％ in EEG classification, while the accuracy of the method based on wavelet is only 88％.That is, the proposed method is feasible for the melancholia diagnosis and research.

关键词： CSSD SVM EEG Melancholia

作者: Jian-Jun Shi Qing-Wu Yuan La-Wu Zhou

作者单位: Hunan Vocational College of Railway Technology Zhuzhou, 412000, China College of Electrical and Information Engineering Hunan University Hunan, 410082, China

会议类型: 国际会议

会议名称: 2010 International Conference on Signal and Information Processing(2010年IEEE信号与信息处理国际会议 ICSIP2010)

会议地点: 长沙

会议语种:英文

页码: 14-18

在线出版日期: 2010-12-14（万方平台首次上网日期，不代表论文的发表时间）